Specifying a repeated measures model using fitrm

Hello,

I'm trying to fit a repeated measures model to a dataset using fitrm, but I'm having a hard time understanding how to specify the model using the help page and examples.

I measured the scalar response FR from 4,894 individuals, termed NEU. I measured FR from each NEU 16 times, while manipulating two variables in a factorial design: CUE (8 categorical values) and STATE (2 categorical values).

Here are the first and last few rows of the table containing the data for clarity:

Using fitrm.m, How can I model the effect of STATE and CUE on FR (FR~CUE*STATE), while accounting for the fact that FR was repeatedly measured from each NEU?

Answer 1

[Sai Pavan]

Hi Matt,

I understand that you are trying to model the effect of “STATE” and “CUE” on “FR” as well as account for the fact that “FR” was repeatedly measured from each “NEU”.

Please try using the "fitrm" function as given below to model the repeated measures:

rm = fitrm(data, 'FR1-FR16~CUE*STATE', 'WithinDesign', [1 4]);

The “FR1-FR16” specifies the repeated measures of “FR”. The “CUE*STATE” specifies the fixed effects of “CUE” and “STATE”, as well as their interaction. By providing the “WithinDesign” and “[1 4]” parameters to the “fitrm” function, we indicate that the repeated measures are within-subjects factors. In this case, the first index (1) represents the column number of repeated measures of “FR”, and the second index (4) represents the column number of the subject identifier “NEU”.

Please refer to the below documentation to learn more about the “WithinDesign” parameter in “fitrm” function: https://www.mathworks.com/help/stats/fitrm.html#:~:text=WithinDesign%20%E2%80%94%20Design%20for%20within%2Dsubject%20factors

Hope it helps.

Regards,

Sai Pavan